Lithium Battery Fire Safety: Suppression and Response
Lithium Battery Fire Safety: Suppression and Response
Lithium battery fires are different from ordinary electrical fires, and treating them the same gets people hurt and property lost. The energy is stored chemically, so a “dead” pack can reignite hours later, and the off-gases are toxic. This guide covers how lithium fires behave, what suppression actually works, the compartmentation and code expectations, and the response plan a procurement manager should require before a pack ships to site.
Why lithium fires behave differently
A lithium-cell thermal runaway releases its own oxygen, so it does not need ambient air to keep burning—classic “remove the oxygen” firefighting fails. Heat from one cell spreads to neighbors in a cascade, and the pack can reach several hundred degrees internally, enough to ignite nearby combustibles. Worse, a quiescent pack can reignite long after it appears out, because residual energy keeps cells cooking. A lithium battery manufacturer builds cells to resist this, but no cell is immune once runaway starts, which is why containment and cooling—not extinction—are the realistic goals, and why the response plan assumes a long, monitored event rather than a quick knock-down.
The thermal runaway cascade
Runaway begins when a cell exceeds its thermal threshold—from overcharge, puncture, external heat or internal defect. It vents flammable electrolyte vapor, heats adjacent cells, and they in turn go into runaway, propagating through the module. The speed depends on cell chemistry and spacing; LFP is more resistant than NMC, and good module spacing slows propagation. A semi-solid-state battery carries less flammable electrolyte, which slows but does not stop the cascade, so the response plan must assume any lithium pack can propagate and plan for cooling the whole bank, not just the first cell. Designing for propagation—not assuming it cannot happen—is the difference between a contained event and a building fire.
Suppression: what works and what doesn’t
| Method | Effect on lithium fire |
|---|---|
| Water (copious) | Cools cells, slows propagation; the primary tool |
| ABC dry powder | Knocks flame, does not cool; reignition likely |
| CO2 | Little effect; no cooling, oxygen returns |
| Class D agent | For metal fires, not the pack fire per se |
| Inert gas (Novec/FM200) | Protects room, limited cell cooling |
The table is blunt: water in large volume is the proven first response because it cools, which is what actually stops propagation. Dry agents hide the flame while the cells keep cooking, so they must be followed by water cooling. Specify a suppression design that delivers water volume to the bank, and reject any plan that relies on powder alone for a lithium install, because the flame-out is not the end of the incident—the cooling is.
Fixed suppression system design
For stationary banks—a home energy storage system or a C&I cabinet—specify early smoke/heat detection and an automatic water-deluge or mist system sized to the bank’s thermal mass, not the room’s. Add gas detection for off-gases (CO, VOC) so a smoldering cell alarms before flame. Pair detection with the BMS so the pack disconnects on alarm. A credible battery application solutions design integrates detection, disconnect and suppression as one system rather than bolting a fire panel onto a battery after the fact, because the interface between BMS and fire system is where response time is won or lost, and seconds decide whether one module or the whole bank is lost.
Module-level containment and barriers
Beyond active suppression, passive compartmentation limits damage. Fire-rated barriers between banks, non-combustible enclosures, and module-level isolation that stops a single cell’s runaway from reaching the next module all reduce the event to a local failure. Spacing, intumescent wraps around modules, and fire-rated room construction are engineering choices that complement water systems. For a containerized ESS, the design question is whether one module can fail without taking the container—if the answer is no, the passive design is inadequate regardless of how good the sprinklers are. Procurement should ask for the propagation test result and the barrier specification together, because they are two halves of one safety case.
Mobile and containerized response
Shipping containers and vehicles need different tactics: isolate the unit, evacuate the radius, and apply water from a distance to cool, never opening the enclosure to “see.” For a containerized ESS, specify external sprinkler coverage and a fire department connection, plus clear placarding of the lithium hazard and the recharge risk. The on-site response card should state plainly that reignition is possible and the pack must be monitored for 24–48 hours after apparent extinguishment. This is not cautious overkill; documented lithium incidents show reignition well beyond the initial event, and the response card is the document that tells the first responder what they are actually facing.
Personal protective and toxic gases
Lithium off-gases include hydrogen fluoride, carbon monoxide and various organics—lethal in enclosed spaces. Responders need SCBA, and occupants need evacuation, not re-entry to “save” the pack. A sodium-ion battery is intrinsically harder to ignite and emits less of these gases, a genuine safety advantage in occupied spaces, but it is not gas-free, so the PPE and ventilation rules still apply. Write the gas hazard into the safety data and the response plan, because the toxic risk kills faster than the flame, and a responder who enters without air pays with their life, not their equipment.
Code and insurance expectations
In the US, NFPA 855 and the International Fire Code set separation, detection and suppression expectations for stationary storage; insurers (FM Global, UL) add their own criteria that often exceed the code minimum. A system that meets the letter of the fire code but fails the insurer’s spacing rule may still be uninsurable, which is a commercial dead end. Confirm both the AHJ requirement and the carrier’s criterion before finalizing the layout, because the gap between “permitted” and “insured” is where many projects stall. A lithium battery manufacturer can supply the propagation data the insurer needs, but the buyer must request it explicitly—it is rarely volunteered.
First-responder training
The best suppression plan fails if the local fire department treats the pack like a wiring fire. Provide the response card, the SDS off-gas section, and a site briefing to the responding authority, and keep a copy at the site entrance. Training covers: do not open the enclosure, apply water from distance, expect reignition, evacuate for toxic gas. For a fleet or multi-site operator, a standard briefings package per site turns a confusing arrival into a competent response. The cost of the briefing is trivial next to a delayed or wrong first action, and it is the buyer’s responsibility to supply it, not the fire department’s to guess.
Emergency response plan checklist
- Detect early via smoke, heat and gas sensors tied to the BMS disconnect.
- Evacuate the radius; lithium fires are not a DIY extinguish.
- Apply copious water to cool; expect prolonged and possible reignition.
- Ventilate the space for toxic off-gas; responders wear SCBA.
- Monitor the pack 24–48 h post-event before declaring safe.
- Post a site-specific response card with the lithium hazard and contacts.
- Brief the local fire authority and keep the SDS on site.
Procurement requirements that save lives
Put fire safety in the RFQ as hard requirements, not suggestions. Demand the propagation test data (UL 9540A or equivalent) showing whether and how fast a cell-to-cell fire spreads in the actual module, because that single test tells you if your site needs room-level suppression or just module isolation. Require the BMS to disconnect on thermal event and to report cell temperatures to the fire system. Ask for the manufacturer’s response card and confirm your local fire authority accepts the suppression design. A buyer who skips this is purchasing a hazard the insurer may later refuse to cover, and the gap shows up exactly when it is most dangerous, after a fire has already started and the plan is untested.
Installation and spacing best practices
Suppression is the backstop; good installation prevents the event. Keep banks out of occupied sleeping spaces and away from escape routes, separate them from combustible storage by the distance your code and insurer specify, and avoid siting them directly above people or critical equipment. Provide a non-combustible, well-drained base and clear access for responders. For a request a quote engagement we lay out the spacing on your actual floor plan rather than quoting a generic rule, because the right separation depends on bank size, barrier rating and the surrounding fire load. A pack installed against code but crammed beside palletized cardboard is not actually safe, and the suppression system cannot compensate for a bad layout—design the space first, then the suppression.
Frequently asked questions
Can you put out a lithium fire with a normal extinguisher? Powder knocks flame but does not cool; water in volume is the real tool because runaway is self-oxygenating.
Why does it reignite? Residual cell energy keeps cooking after the flame dies, so monitor the pack for 24–48 hours.
Is sodium safer in a fire? Harder to ignite and lower toxic off-gas, but PPE and ventilation still apply.
What test proves my bank is safe? UL 9540A propagation testing on the actual module shows whether fire spreads cell to cell.
Written by Karl at China Battery Technology. Send us your site layout and we will design detection, disconnect and suppression—request a quote for a fire-safe battery plan.
Related Reading
- battery application solutions hub
- MSDS for Lithium Battery: What Importers Must File
- CE vs UL Marking: Which Your Product Actually Needs
